Scalable Link Prediction in Dynamic Networks via Non-Negative Matrix Factorization

TL;DR

This paper introduces a scalable temporal latent space model using non-negative matrix factorization for link prediction in dynamic networks, effectively capturing evolving user interactions over time.

Contribution

It presents a novel scalable model with multiple optimization algorithms for dynamic link prediction, outperforming existing methods in accuracy and scalability.

Findings

Significantly outperforms existing approaches in real-world networks.

Demonstrates high scalability with incremental update algorithms.

Achieves quadratic convergence rate in optimization.

Abstract

We propose a scalable temporal latent space model for link prediction in dynamic social networks, where the goal is to predict links over time based on a sequence of previous graph snapshots. The model assumes that each user lies in an unobserved latent space and interactions are more likely to form between similar users in the latent space representation. In addition, the model allows each user to gradually move its position in the latent space as the network structure evolves over time. We present a global optimization algorithm to effectively infer the temporal latent space, with a quadratic convergence rate. Two alternative optimization algorithms with local and incremental updates are also proposed, allowing the model to scale to larger networks without compromising prediction accuracy. Empirically, we demonstrate that our model, when evaluated on a number of real-world dynamic networks, significantly outperforms existing approaches for temporal link prediction in terms of both scalability and predictive power.